Lesson 6 of 10
Autonomous Maintenance
Five lessons of measuring is enough. This is the first pillar, and the first one that changes anything — not by buying equipment or hiring technicians, but by moving a small amount of responsibility to the person already standing at the machine.
Step one
Read the material
Three ideas. Why cleaning is not about cleanliness, the difference between wear you caused and wear you did not, and the real cost of the person who notices not being the person who can act.
Concept one
Cleaning is inspection
Autonomous Maintenance almost always starts with operators cleaning their own equipment, and almost everyone misunderstands why. It is not a tidiness programme. It is not about looking professional for visitors.
You cannot see a fault on a dirty machine. A weeping seal looks like every other oily patch. A hairline crack disappears under grime. A fastener backing out looks the same as one that is tight. Coolant leaking where it should not be goes unnoticed because there is coolant everywhere.
Cleaning is the only activity that reliably puts a human hand and eye on every surface of the machine on a schedule. The dirt is not the problem; the dirt is what has been hiding the problem. Plants that begin this properly are always startled by the first month, when a machine everybody considered fine produces a list of thirty abnormalities.
That list is the point. It is called finding abnormalities, and it is deliberately a low bar — anything not as it should be. A loose guard. A gauge nobody can read. A missing bolt. A noise that is new. The operator does not have to diagnose it or fix it. They have to notice it and tag it, which is a skill, and it is teachable.
The formal method is seven steps, and cleaning is only the first: initial clean, eliminate the sources of contamination, set cleaning and lubrication standards, general inspection, autonomous inspection, standardise, then full self-management. Most plants that fail at AM fail because they did step one and called it done.
Concept two
Most wear is not wear
Equipment deteriorates two ways, and telling them apart changes what you do about it.
Natural deterioration is the wear a machine suffers from being used correctly. Bearings have a life. Cutting edges dull. Belts stretch. This is real, it is predictable, and it is what planned maintenance exists to manage — that is lesson seven.
Forced deterioration is the wear you caused. Running without lubrication. Contamination getting into places it was never meant to reach. Fasteners left loose so a machine shakes itself apart. Operating outside the conditions the equipment was designed for. None of this had to happen.
Here is the uncomfortable part: in most plants that have never done AM, the majority of breakdowns are forced deterioration, not natural. The bearing did not fail because bearings fail. It failed because grit got past a seal nobody had cleaned, or because it was never greased on the interval, or because it had been running out of alignment for months and nobody heard it.
This is why AM produces results that look impossible. A plant is not making its machinery better than new; it is stopping doing the things that were destroying it. And that work cannot be done by a maintenance department on a schedule, because it is not periodic — it is continuous, and it happens at the machine, where only the operator is standing.
Concept three
The person who notices is not the person allowed to act
Picture a five-minute job. A guard needs re-seating, a sensor needs wiping, a fitting needs a quarter turn. The operator sees it immediately and could do it in five minutes.
Instead they raise a call. It queues. A technician arrives forty, ninety, two hundred minutes later, does five minutes of work, and leaves. The repair took five minutes. The stop took two hours. Nearly all of that loss was waiting, and waiting is not a maintenance problem — it is an organisational one.
Autonomous Maintenance closes that gap by moving a defined set of tasks to the operator: cleaning, lubrication, fastener checks, routine inspection, simple adjustments. The technician's queue empties of trivia, which means the technician is finally free to do the skilled work that actually needs them.
What AM is not: it is not operators doing repairs, and it is not a way to reduce the maintenance headcount. Both misreadings kill the programme, and the second one kills it fastest — the moment operators believe AM is a redundancy exercise, the abnormality tags stop appearing.
The boundary has to be written down and it has to be honest. Some tasks require isolation, a qualification, or an authority the operator does not hold, and those stay with maintenance permanently. Drawing that line clearly is what makes operators willing to own everything on their side of it.
Step two
Work it with your hands on it
One machine, one week. The last slider moves routine tasks from the technician to the operator — and what changes is not the work itself, but who is standing there when it needs doing.
Hands on it, off the screen. Take one week of maintenance call records for a single machine. For each call, write two numbers: how long the technician was actually working, and how long the machine sat between the call being raised and the technician arriving.
Total both columns. In most plants the second is several times the first, and nobody has ever put the two side by side. That comparison is the entire business case for Autonomous Maintenance, and it takes an afternoon to produce.
Step three
Prove you understood it
One question per idea, new numbers every time.
All three solid.
You know why cleaning is inspection, why most wear was avoidable, and where the real cost of a small fault actually sits. Lesson seven takes the other half — Planned Maintenance, and the two numbers that decide availability: mean time between failures, and mean time to repair.
Modern Total Productive Maintenance (TPM)
Lesson six of ten. This class is taught from the Total Productive Maintenance series — the eight pillars, the six big losses, and the roadmap into AI and predictive manufacturing. Written by our founder, Dr. Gene A Constant, and donated to the Foundation.
Read on Kindle The whole class AI + TPMGSU is a 501(c)(3) nonprofit and book sales are its only income. A purchase is not a tax-deductible donation. As an Amazon Associate, Global Sovereign University earns from qualifying purchases. This lesson is free and always will be. Global Sovereign University: Different by Design. Better by Mission.